TopolearnDefault ticks are often fine but wrong for percentages, currency or categories; precise control is what makes charts publication-ready.
Limits come from the data plus a small margin. Override them with ax.set_xlim(left, right) and ax.set_ylim(bottom, top); pass None for one side to keep it automatic. Reversing the order, set_ylim(10, 0), flips the axis. ax.margins(0) removes the padding.
Ticks are chosen by a locator and labelled by a formatter, both attached to each Axis (ax.xaxis, ax.yaxis).
Quick, explicit control:
ax.set_xticks([0, 5, 10], labels=["start", "middle", "end"])
ax.tick_params(axis="x", rotation=45, labelsize=8)Rule-based control, which keeps working when data changes:
ticker.MultipleLocator(5): a tick every 5 units.ticker.MaxNLocator(nbins=4, integer=True): at most a few integer ticks.ticker.PercentFormatter(xmax=1): shows 0.25 as 25%.ticker.StrMethodFormatter("{x:,.0f}"): thousands separators.ticker.FuncFormatter(lambda v, pos: ...): anything you can compute.
Set them with ax.yaxis.set_major_locator(...) and ax.yaxis.set_major_formatter(...). Passing a format string directly, ax.yaxis.set_major_formatter("{x:.1f} s"), creates a StrMethodFormatter for you.
Minor ticks are off by default for linear axes; ax.minorticks_on() adds them. A light grid, ax.grid(True, alpha=0.3), helps readers look up values.
Hard-coded tick labels via set_xticks(labels=...) are tied to fixed positions. If the limits change later, they will no longer match; prefer a formatter when the labels are derived from values.
Example
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
from matplotlib import ticker
fig, ax = plt.subplots()
ax.plot([0, 1, 2, 3], [0.12, 0.35, 0.61, 0.9])
ax.set_xlim(0, 3)
ax.set_ylim(0, 1)
ax.xaxis.set_major_locator(ticker.MultipleLocator(1))
ax.yaxis.set_major_formatter(ticker.PercentFormatter(xmax=1))
ax.grid(True, alpha=0.3)
fig.canvas.draw() # ticks are computed at draw time
print([t.get_text() for t in ax.get_yticklabels()][:3]) # ['0%', '20%', '40%']
fig.savefig("ticks.png")Remember
set_xlim/set_ylimset ranges; reversed limits flip an axis- Locators choose tick positions, formatters choose their labels
- Prefer formatters over hard-coded labels when labels derive from values
Watch out for
- Setting fixed tick labels and then changing limits, leaving labels that no longer match positions.